New access points discover a Catalyst 9800, start the join, and stall while transferring the image. A small ping to the controller works, and a large ping across the same path fails. What is the fault?
Select an answer to reveal the explanation.
Short Explanation
Small packets arrive and big ones do not. Your access points can find the controller and then choke on the image, which is a big transfer. The path has to carry that size, and the guest name is not the choke.
Full Explanation
Discovery and the start of the join use smaller exchanges. The image transfer uses larger packets. A path that passes a small ping and drops a large one will let the access point find the controller and then stall on the image. The fault is that path, often a tunnel or a device with a too-small MTU and no working fragmentation. The guest SSID is unrelated to an access point downloading its image. Antenna gain does not decide whether a large ping works. The beacon interval does not fragment an image and is not compared with a ping size. Discovery can succeed on small packets while the image transfer stalls on large ones. The path has to pass the size the transfer actually uses. Exam caveat: fixing the MTU on one hop is not enough if the next hop still drops the large packet. Test the large ping from the access-point VLAN all the way to the controller. Operational check: after the path passes a large ping, restart one access point and confirm the image transfer finishes and the access point joins.